pychrono and multicore support

[Sean You]

Hi All,

I'm working on simulating a system in pychrono with a moderate number of particles (~300) - The runtime seems to be relatively slow - I was wondering if there was any support for using the multicore solvers in pychrono/any other recommended performance improvements or do I need to port my simulation over to C++?

Best,

Sean

[Dan Negrut]

Sean – what would help is to make sure that the meshes used for collision for the ~300 is very simple. Better yet, you use simple geometries, primitives, that is. That would help.

 

Also, going to C++ would help too. But keeping the collision meshes simple is where I would start.

 

Good luck.

Dan

---------------------------------------------

Bernard A. and Frances M. Weideman Professor

NVIDIA CUDA Fellow

Department of Mechanical Engineering

Department of Computer Science

University of Wisconsin - Madison

4150ME, 1513 University Avenue

Madison, WI 53706-1572

608 772 0914

http://sbel.wisc.edu/

http://projectchrono.org/

---------------------------------------------

--
You received this message because you are subscribed to the Google Groups "ProjectChrono" group.
To unsubscribe from this group and stop receiving emails from it, send an email to projectchron...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/projectchrono/05690b49-0042-4508-98a7-225879871a24n%40googlegroups.com.

[Radu Serban]

Hi Sean,

 

Performance of code based on PyChrono is only marginally (if at all) worse that that of code directly using the Chrono C++ libraries. PyChrono is just a set of SWIG-generated wrappers around the C++ Chrono code.

 

However, if you think that your code would benefit from using the Chrono::Multicore module (note that this is not a guarantee, as 300 particles is not all that many – you may want to provide some details on what geometry you use for these bodies), you will have to port your code to C++ because the Chrono::Multicore module is currently *not* SWIG-wrapped and as such not available through PyChrono.

 

Best,
Radu

 

From: projec...@googlegroups.com <projec...@googlegroups.com> On Behalf Of Sean You
Sent: Tuesday, March 25, 2025 9:51 PM
To: ProjectChrono <projec...@googlegroups.com>
Subject: [chrono] pychrono and multicore support

 

Hi All,

--

[Sean You]

I'm currently simulating these particles as spheres with some amount of friction and adhesion - the only complex geometry is the container for these particles which is attached to a linear displacement motor (to agitate them etc). Are there any additional ways in which I could try to speed up the simulation/further simplify the particle geometry?

In a related question would it make sense to use the granular module in C++ as I increase the number of particles? Ultimately I'm interested in looking at 10k+ , and looking at if they can remain well separated (ie acting as single particles). In the physical system I'm modeling the container is on the order of centimeter scale, and the particles would have micron to millimeter diameters. 

Best,

Sean